Persons with NIDDM have abnormal oxygen consumption (VO2) responses to exercise even in the absence of clinical cardiovascular disease. We hypothesize that in persons with NIDDM, compared to nondiabetic persons, decreased maximal and dynamic cardiac output and/or A-VO2 responses are associated with the observed impaired VO2 responses. We also hypothesize that the decreased VO2 observed in NIDDM is associated with decreased levels of habitual physical activity level compared to control values. In addition, we hypothesize that exercise training will improve cardiovascular exercise performance and habitual physical activity level. Subjects will include 20 sedentary persons with uncomplicated NIDDM, 20 healthy persons of similar weight, age and activity level and 20 lean persons of similar age and activity level as the persons with NIDDM. We will measure the dynamic and maximal responses of VO2 cardiac output and A-VO2 difference (by direct Fick and hemodilution methods) during exercise in persons with NIDDM. Slowed maximal responses of cardiac output would suggest an impairment in oxygen delivery in NIDDM. Additionally, or alternatively, A-VO2 difference responses may be also be abnormal suggesting a peripheral impairment. To determine whether the decreased VO2max response observed in NIDDM is associated with decreased habitual physical activity level we will compare the three groups using doubly-labeled water techniques. A four month exercise training program will follow the testing procedures with the tests to be repeated upon completion. The importance of these studies is 1) to evaluate the underlying causes of the apparent cardiovascular abnormalities observed in uncomplicated NIDDM; 2) to evaluate the functional significance of these findings, and 3) to evaluate the modifiability of the responses. This project will enable further understanding of the effects of diabetes on cardiovascular function during exercise as well as the functional significance of exercise impairment in diabetes.